OSDev.org

i'm beginning to think that zaleschiemilgabriel is a OS code copier (someone who copies code and does not have a clue nor wants to learn what it does). Also with a nick that's longer than the list of issues Windows has, I would say no one is/will be interested in what he has to say.

A language is a language, get over assembly for god's sake.

i seeked for code from zalahjflezzvugabriel, and the surprise...
none of his post contain code... am i blind?

How good a OS is, is not down to the language that the OS was written in, but the mind or minds of the coder or coder's that written it.

It does not matter what languages that are used. What matters is the end result; how stable, portable, extendable, and ease of use the system is depending on the requirements specification used and what is needed.

Real programmers use and know what tools they need to use to build the software, whether it be assembly, C, C++, etc.

zaleschiemilgabriel wrote:Real OS-es are written in assembly.

Just a couple questions; How would you achieve a high level of portability in a kernel built in assembly? And why would you do that much with an assembler as ugly as intel x86? Something like M68k i would understand, but not x86. To quote the AMD CEO "there's nothing wrong with x86, it just doesn't make a lot of sense."

Secondly, If you're using a low level language, C for example, your standard compiler can do optimizations you would not do yourself(its hard to explain w/o pictures, but I'm pretty sure M$ has a ton of docs on how their optimization works for example). It stems from unless you have an intimate knowledge of the CPU there's an order of routines which would work faster then what you'll write, and it kind of goes along with how adding well placed NOPs can actually improve execution.

Which is exactly why people buy ICC (the Intel compiler set): It produces better assembly code than assembly coders can.

Yayyak wrote:Which is exactly why people buy ICC (the Intel compiler set): It produces better assembly code than assembly coders can.

Ah, I love it when math and logic can outdo humans. It smells like victory.

Alboin wrote:

Yayyak wrote:Which is exactly why people buy ICC (the Intel compiler set): It produces better assembly code than assembly coders can.

Ah, I love it when math and logic can outdo humans. It smells like victory.

But I bet an assembly programmer can still take ICC's output assembly code and optimize it further than the compiler could. Besides, compilers are made by humans, anyways.

lollynoob wrote:

Alboin wrote:

Yayyak wrote:Which is exactly why people buy ICC (the Intel compiler set): It produces better assembly code than assembly coders can.

Ah, I love it when math and logic can outdo humans. It smells like victory.

But I bet an assembly programmer can still take ICC's output assembly code and optimize it further than the compiler could. Besides, compilers are made by humans, anyways.

I doubt it. That particular compiler was created using information only available to the designers of intel's chips. As it was created by the chip designers, and the chip designs are not public, it is therefore logical that they have more information than you can ever have, and therefore also logical that you can never outdo them.

To be perfectly honest, I doubt if most "assembly programmers" even know how pipelining affects performance. Most members of this board are omitted from that description, however!

You guys suck! You need me to post code just to prove something? That's just another side-effect of C programming. There are two main reasons why people post code: they post assembly code because they don't know assembly and they **** it up big time and they post C code because their code gets too bloated to keep track of. In both cases, they need help from the experts. I do not post code because I know how to code. It's OS design that I'm unfamiliar with!
At least I had the guts to start coding by myself, and I did something good without help and stupid questions: PCI enumeration, ACPI tables enumeration, VESA video modes enumeration/initialization (and I am now working on ATA controller driver). I did/redid all those thing in assembly. I just don't know how to put them together (basically it's theory that I'm after; the practical part - coding/debugging/testing - I'm all to acquainted with)!

they post assembly code because they don't know assembly

you are a real troll!

postng assembly code is a way to contribute to help beginers.. and to have suggestions from abstraction specialists, that will orient your design to link all your little code together.

zaleschiemilgabriel wrote:You guys suck! You need me to post code just to prove something? That's just another side-effect of C programming. There are two main reasons why people post code: they post assembly code because they don't know assembly and they **** it up big time and they post C code because their code gets too bloated to keep track of. In both cases, they need help from the experts. I do not post code because I know how to code. It's OS design that I'm unfamiliar with!
At least I had the guts to start coding by myself, and I did something good without help and stupid questions: PCI enumeration, ACPI tables enumeration, VESA video modes enumeration/initialization (and I am now working on ATA controller driver). I did/redid all those thing in assembly. I just don't know how to put them together (basically it's theory that I'm after; the practical part - coding/debugging/testing - I'm all to acquainted with)!

From your manner of speaking and what you've said, It's entirely probable that your code and design is, to coin a technical phrase, a bag of shite.

As to your other comments, I won't dignify them with the point-by-point rebuttal I usually do. You obviously like the attention.

Cheers,

James

code does not influence the design! it's the other way around! if you think that way, you probably use someone else's code (to get you started coding the rest...)

zaleschiemilgabriel wrote:code does not influence the design! it's the other way around! if you think that way, you probably use someone else's code (to get you started coding the rest...)

If you reread my post, you will find no mention or implication that implementation influences design.

Learn to read.

James, I respect you so I do not expect anything from you. I understand that you think yourself superior because you wrote a tutorial that helps lots of people here, and you are (the tutorial helped me a lot too), but I want to do things a bit different, not be a copy-cat.
Also I think that what you said before about me instigating everybody also applies to you. If you keep calling me a troll, that's what I'll become!